Opal is a free, no-code AI app builder from Google Labs that turns a plain-language description into a shareable mini-app. PR teams use Opal to build small internal tools, a media monitoring dashboard or a client reporting workflow, without writing code or waiting on an engineering team.
What is Opal?
Opal launched as a US-only public beta in late July 2025, according to Google's launch post, letting users describe an app in natural language and refine it on a visual canvas where each step, a prompt, a model call or a tool, chains into the next. Google expanded Opal to more than 160 countries by November 2025 and integrated it into the Gemini web app, according to a September 2026 product overview.
Opal sits below full app builders that generate a deployable codebase. Its output is a shareable workflow other people can run by link, not a hosted application with a backend, which is what makes it fast to turn a repeated prompt sequence into something a colleague can use without an API integration.
Why does Opal matter for PR teams specifically?
Opal matters for PR teams because most of the tools a communications department wants, a coverage tracker, a statement-approval router, a press-list formatter, are exactly the kind of small, single-purpose apps Opal is built to produce, and none of them previously justified pulling an engineer off a product roadmap.
Why it works: Opal added an Agent step on February 24, 2026, powered by Gemini 3 Flash, which plans an approach, selects its own tools and models, and executes rather than following a fixed chain a person defined step by step, according to the same product overview. That shift means a comms staffer can describe an outcome, "pull today's coverage, summarize sentiment, and draft a one-line update for the team," and Opal decides how to sequence the research, summarization and drafting steps itself, instead of the person having to wire each step manually.
How are comms teams using Opal today?
Comms teams are using Opal to prototype the internal tools that never made it onto an engineering backlog: a research summarizer that reads multiple source documents and writes bullet points, a brand-asset generator that produces marketing copy and captions from one idea, and a quiz or interactive explainer for a campaign microsite, all examples Google itself lists as example Opal use cases. That puts it alongside the other Google Labs and Workspace AI tools comms teams are already testing, including the image generation covered in EPR's Nano Banana for PR guide and the video generation covered in EPR's Google Vids for PR guide.
Illustrative scenario: a PR team fielding daily inbound press questions during a product launch builds an Opal that takes a reporter's question, checks it against an approved Q&A document, and drafts a response for a staffer to review before sending, cutting the first-draft step from a manual document search to a single link the whole team can open and run.
What technology is Opal actually built on?
Opal is built on Breadboard, an open-source framework Google uses to let a person or an AI agent chain prompts, model calls and tools into a visual graph, which is why an Opal app appears on screen as a series of connected steps rather than as a single chatbot conversation.
Why it works: Because each step in an Opal graph is a discrete, inspectable node, a person building an app can open any single step, see exactly what prompt or tool call runs there, and edit that one step without breaking the rest of the chain, according to the product overview of Opal's architecture. That inspectability is the specific feature that separates Opal from a single long prompt pasted into a chatbot: a comms staffer who notices the summarization step of a media-monitoring app is missing a source can fix that one node without rebuilding the whole workflow from scratch.
What does a second concrete PR workflow look like in practice?
A second concrete PR workflow already in use is a press-list formatter: a comms staffer pastes a raw list of journalist names, outlets and beats copied from a spreadsheet or a media database export, and an Opal app reformats it into the agency's standard pitch-tracking template, flags any row missing an email address, and outputs a file ready to drop into the team's existing outreach tool.
That kind of formatting task is exactly the repetitive, rules-based work a comms coordinator previously did by hand every time a new list came in from a client or a research vendor, and it is also exactly the kind of task Opal's example gallery highlights as a template other users can copy and adapt, rather than building from a blank canvas each time a similar need comes up.
Because an Opal app can be shared by link, a team that builds one useful formatter or tracker can hand it to every account team in the agency without each team needing to describe and rebuild the same workflow independently, which is a meaningfully different distribution model than a script one person keeps on their own laptop.
What are the real limits of Opal?
Opal has no paid tier and no published service-level agreement as of September 2026, according to the product overview cited above, which makes it a fit for prototyping and internal workflows rather than a client-facing or mission-critical system.
Opal also has no separate public API, so an Opal a team builds cannot be embedded directly into another company system; it runs as a standalone shareable app inside Google's own infrastructure, a limit worth checking against IT policy before a comms team routes anything containing client or embargoed information through it.
Because Opal runs inside a person's existing Google account, adoption inside a PR team does not require a separate procurement or vendor-review process the way a new SaaS platform would, which is part of why comms staffers have been able to start experimenting with it directly rather than waiting on an IT-approved tool list.
How does Opal compare to hiring a developer or using a workflow tool?
| Option | When to use it | Trade-off |
|---|
| Opal | Fast prototype of a single-purpose internal tool, no budget for engineering time | Free, but no SLA, no paid tier and no embedding API as of September 2026 |
| Workflow automation tool (e.g. Zapier, Make) | Connecting existing SaaS tools the team already pays for | Requires per-step configuration rather than a plain-language description |
| In-house or contract developer | A tool that needs to embed in another system or handle sensitive client data | Slower to ship, higher cost, but comes with support and security review |
Opal also lowers the barrier for a PR team to test an idea before deciding whether it deserves real engineering investment. A tool built in an afternoon and used by three people for two weeks tells a comms leader whether a coverage tracker or an approval router is worth a formal build, without spending a sprint of engineering time to find out.
CONCLUSION
Opal turns the class of small, single-purpose tool that PR teams have always wanted and rarely gotten, a tracker, a router, a summarizer, into something a non-technical staffer can build directly, and the February 2026 Agent step means the tool now plans its own steps rather than needing every one specified by the person building it. The trade-off is a free, unsupported Google Labs experiment with no embedding API, so it fits internal prototyping today, not a client-facing system handling sensitive data.
